Suresh Duddi wrote: > > Username was NULL. This was the first time I used singlesignon with > the current code base. > > went to netcenter and webmail. On initial login, it asked me if I > wanted to save my username and password in the shell window (really > cool). I said yes. It created signons.txt and then I quit. > > next time I started the viewer, I went to the same place but saw my > username not filled in while my password was. I typed in my username > and clicked on submit. It crashed. > > dp > > si_CheckForUser(char * 0x01306120, char * 0x00000000) line 519 + 9 > bytes > si_OkToSave(MWContext_ * 0x011bf7e0, char * 0x01306120, char * > 0x00000000) line 903 + 13 bytes > SI_RememberSignonData(char * 0x01306120, LO_FormSubmitData_struct * > 0x0012f988) line 1954 + 26 bytes > nsNetlibService::SI_RememberSignonData(nsNetlibService * const > 0x011debb0, char * 0x01306120, LO_FormSubmitData_struct * 0x0012f988) > line 694 + 13 bytes > nsFormFrame::ProcessAsURLEncoded(int 1, nsString & {...}, > nsIFormControlFrame * 0x012dc6cc) line 682 + 23 bytes > nsFormFrame::OnSubmit(nsFormFrame * const 0x012d4944, nsIPresContext * > 0x012bfd20, nsIFrame * 0x012dc670) line 497 > nsImageControlFrame::MouseClicked(nsIPresContext * 0x012bfd20) line > 362 > nsImageControlFrame::HandleEvent(nsImageControlFrame * const > 0x012dc670, nsIPresContext & {...}, nsGUIEvent * 0x0012fdb4, > nsEventStatus & nsEventStatus_eIgnore) line 237 > PresShell::HandleEvent(PresShell * const 0x012537a4, nsIView * > 0x012dc980, nsGUIEvent * 0x0012fdb4, nsEventStatus & > nsEventStatus_eIgnore) line 1431 + 33 bytes > nsView::HandleEvent(nsView * const 0x012dc980, nsGUIEvent * > 0x0012fdb4, unsigned int 8, nsEventStatus & nsEventStatus_eIgnore) > line 789 > nsView::HandleEvent(nsView * const 0x0127b7e0, nsGUIEvent * > 0x0012fdb4, unsigned int 8, nsEventStatus & nsEventStatus_eIgnore) > line 772 > nsView::HandleEvent(nsView * const 0x0127ba70, nsGUIEvent * > 0x0012fdb4, unsigned int 28, nsEventStatus & nsEventStatus_eIgnore) > line 772 > nsViewManager::DispatchEvent(nsViewManager * const 0x01253f40, > nsGUIEvent * 0x0012fdb4, nsEventStatus & nsEventStatus_eIgnore) line > 823 > HandleEvent(nsGUIEvent * 0x0012fdb4) line 64 > nsWindow::DispatchEvent(nsWindow * const 0x0127b130, nsGUIEvent * > 0x0012fdb4, nsEventStatus & nsEventStatus_eIgnore) line 355 + 10 bytes > > nsWindow::DispatchWindowEvent(nsGUIEvent * 0x0012fdb4) line 371 > nsWindow::DispatchMouseEvent(unsigned int 301, nsPoint * 0x00000000) > line 1849 + 15 bytes > ChildWindow::DispatchMouseEvent(unsigned int 301, nsPoint * > 0x00000000) line 1998 > nsWindow::ProcessMessage(unsigned int 514, unsigned int 0, long > 16384282, long * 0x0012fef4) line 1388 + 24 bytes > nsWindow::WindowProc(void * 0x001d05d8, unsigned int 514, unsigned int > 0, long 16384282) line 414 + 27 bytes > USER32! 77e71250()

Setting all current Open/Normal to M4.

per leger, assigning QA contacts to all open bugs without QA contacts according to list at http://bugzilla.mozilla.org/describecomponents.cgi?product=Browser

Part of this report (the crash part) is a duplicate of bug report 1729. The remaining part (about only the password being filled) in is not a bug but rather a quirk of the netcenter login form. Rather than being a single form, it consists of one form for the username and a second form for the password. When a username is entered, it is reflected into a hidden field in the password form. And it is the password form that is submitted to netcenter. So the password is what gets saved and restored. The user name is not on the submitted form and so cannot be saved. It was apparent when single signon was designed that certain sites that used unorthodox forms would not be save-able. But it was clear that such sites would be in the minority and so if the heuristic worked most of the time it would be good enough. It's unfortunate to discover that netcenter is one of those unorthodox sites. In any case, this part of the bug report is being considered as INVALID. Since the bug report has two parts, I have a choice now as to whether to close it out as being a DUPLICATE or as being INVALID. I'll choose the former. *** This bug has been marked as a duplicate of 1729 ***

verified